JP6080726B2 - Moving picture encoding apparatus, intra prediction mode determination method, and program - Google Patents

Description

  The present invention relates to a moving image encoding technique.

  In a conventional video encoding apparatus, in order to determine an intra prediction mode (intra prediction mode), a prediction image is generated for each of a large number of intra prediction modes, and a screen in which a difference between the prediction image and the input image is small It is common to select an intra prediction mode (for example, Non-Patent Document 1).

ITU-T Recommendation H. 265

  In the conventional moving image encoding apparatus, a prediction image is generated for each of a large number of intra-screen prediction modes, so the scale of the circuit for generating the prediction image is increased, and the calculation time for generating the prediction image is increased. There is a problem of doing it.

  One of the main objects of the present invention is to solve the above-described problems, and it is possible to simplify the determination of the intra prediction mode and to realize a configuration that can suppress an increase in circuit scale and an increase in calculation time. The main purpose.

A moving image encoding apparatus according to the present invention includes:
A video encoding device that performs inter-frame prediction and intra-frame prediction,
For each image block divided from the image frames that make up a moving image, input the motion vector calculated for inter-frame prediction, analyze the input motion vector, and perform multiple intra-screen predictions for each image block A video encoding apparatus comprising: an intra-screen prediction mode determining unit that selects an intra-screen prediction mode to be applied to intra-screen prediction from among the modes.

  In the present invention, by using a motion vector calculated for inter-frame prediction, determination of the intra prediction mode can be simplified, and an increase in circuit scale and calculation time can be suppressed. .

FIG. 3 is a diagram illustrating a configuration example of a moving image encoding apparatus according to Embodiment 1. H. The figure explaining the intra-screen prediction mode of H.265. The flowchart figure which shows the determination procedure of the prediction mode in a screen which concerns on Embodiment 1. FIG. FIG. 6 is a diagram illustrating a configuration example of a moving image coding apparatus according to Embodiment 2. The flowchart figure which shows the determination procedure of the prediction mode in a screen which concerns on Embodiment 2. FIG. FIG. 10 is a diagram illustrating a configuration example of a moving image coding apparatus according to Embodiment 3. The flowchart figure which shows the determination procedure of the prediction mode in a screen which concerns on Embodiment 3. FIG. The figure which shows the hardware structural example of the moving image encoder which concerns on Embodiment 1-3.

  In the first to third embodiments, a description will be given of a video encoding apparatus that simplifies the determination of the intra prediction mode and suppresses an increase in circuit scale and an increase in computation time while suppressing deterioration in image quality.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram showing a moving picture encoding apparatus 100 according to the present embodiment.

In FIG. 1, an intra-screen prediction mode determination unit 101 receives a motion vector of an encoding target block from the motion compensation prediction unit 104, and inputs a setting parameter from the outside, and evaluates the motion vector.
Then, the intra-screen prediction mode determination unit 101 determines the intra-screen prediction direction based on the evaluated information and the setting parameter, and outputs the prediction direction to the intra-screen prediction image generation unit 102.
Note that the encoding target block is an image block divided from image frames constituting a moving image.

The intra-screen prediction image generation unit 102 receives the notification of the intra-screen prediction mode determined by the intra-screen prediction mode determination unit 101, and obtains local decoding pixels around the encoding target block necessary for prediction image generation from the local decoding image memory 103. input.
Then, the intra-screen prediction image generation unit 102 generates and outputs an intra-screen prediction image corresponding to the intra-screen prediction mode.

The local decoded image memory 103 receives the peripheral image of the encoding target block from the adder 106 and the intra prediction image generation unit 102 and accumulates pixels necessary for generating the intra prediction image.
Then, the local decoded image memory 103 outputs the pixel to the intra-screen prediction image generation unit 102.

The motion compensation prediction unit 104 receives the original image of the encoding target block and an image to be referred from the frame memory 105, compares the encoding target block with the reference image, and frames the reference image with the highest similarity. It is taken out from the memory 105 and outputted to the selector as an inter-frame prediction image.
Then, the motion compensation prediction unit 104 outputs the motion vector of the position information with the highest similarity of the reference image to the intra prediction mode determination unit 101 and the encoding control unit 111.

The frame memory 105 receives the local decoded image subjected to the deblocking filter process from the deblocking filter unit 112 and accumulates pixels necessary for generating the inter-frame prediction image.
Then, the frame memory 105 outputs pixels necessary for generating the inter-frame prediction image to the motion compensation prediction unit 104 based on the position information based on the motion vector input from the motion compensation prediction unit 104.

  The adder 106 inputs the pixel difference value obtained by local decoding of the encoding target block from the inverse frequency transform inverse quantization unit 109 and the prediction image from the selector 113, adds them, and outputs the addition result to the deblocking filter unit 112 and the local decoded image. Output to the memory 103.

  The subtractor 107 receives and subtracts the encoding target block and the prediction image of the encoding target block, and outputs the subtraction result (pixel difference value) to the frequency transform quantization unit 108.

The frequency conversion quantization unit 108 receives the pixel difference value from the subtractor 107, performs frequency conversion calculation such as direct conversion and discrete cosine conversion (DCT), and then quantizes the calculation result.
Then, the frequency transform quantization unit 108 outputs the quantized frequency transform coefficient to the entropy coding unit 110 and the inverse frequency transform inverse quantization unit 109.

  The inverse frequency transform inverse quantizer 109 receives the quantized frequency transform coefficient from the frequency transform quantizer 108, performs inverse quantization on the coefficient, performs inverse frequency transform operation, and decodes it. The pixel difference value is output to the adder 106.

  The entropy encoding unit 110 receives the frequency conversion coefficient after quantization, performs entropy encoding, and outputs it as an encoded stream.

  The encoding control unit 111 controls the entire encoding.

  The deblocking filter unit 112 performs a deblocking filter operation on the locally decoded image input from the adder 106 and outputs the result to the frame memory 105.

1 is a circuit, for example.
Further, the “˜part” can be realized by a program.
When “˜unit” is realized by a program, it is conceivable that the moving image encoding apparatus 100 has a hardware configuration shown in FIG. 8, for example.
That is, a program that realizes the function of “˜unit” is stored in the external storage device 902, the program of “˜unit” is loaded into the main storage device 903, and the arithmetic device 901 executes the loaded program. .

Hereinafter, an operation according to the present embodiment will be described.
Since the moving picture coding apparatus 100 according to the present embodiment has main features in the intra prediction mode determination unit 101, the operation of the intra prediction mode determination unit 101 will be described below.

The intra-screen prediction mode determination unit 101 inputs parameters for setting and a motion vector derived from the motion compensation prediction unit 104 for inter-frame prediction of the encoding target block.
Specifically, the setting parameters are a threshold value for vector length evaluation and a threshold value for vector gradient evaluation.
A motion vector equivalent to inter-frame prediction is derived even in a picture that is predicted in the screen without using a reference image between frames, and the motion vector is input to the intra-screen prediction mode determination unit 101 for determining the prediction mode in the screen.

For example, H.M. 265 DC prediction mode in intra prediction (hereinafter, simply referred to as DC prediction) of 34 directional prediction mode (hereinafter, simply referred to as directional prediction), including the Plan a r prediction mode (hereinafter, simply Plan referred to as a r prediction) it has the option of (Figure 2).
Since the direction of the arrow in FIG. 2 indicates the direction in which the reference image is referenced, intra prediction is performed in the direction opposite to the arrow.
In general, DC prediction is suitable for prediction of a flat texture region having no direction and no edge, and Planar prediction is used for prediction of a flat texture with little change in pixel values such as gradation. Is suitable.

  The intra-screen prediction mode determination unit 101 according to the present embodiment determines an intra-screen prediction mode suitable for a block to be encoded by evaluating a motion vector using a parameter.

First, evaluation of a motion vector with respect to a threshold for vector length evaluation will be described.
The length of the motion vector represents the perspective of the position of the reference image in inter-screen prediction.
That is, when the motion vector is long, the inter-frame reference image is at a position away from the encoding target block, and when it is short, the inter-frame reference image is at a position close to the encoding target block.
When the motion vector is long, the encoding target is a block having a certain amount of motion in a certain direction, and it is highly likely that directional prediction other than DC prediction is suitable for intra prediction.
If the motion vector is short in inverse coding object is a block not too have a motion, it is likely that the facing is DC prediction and Plan a r prediction having no direction among directions prediction.

Next, the motion vector evaluation with respect to the vector inclination evaluation threshold will be described.
The inclination of the motion vector represents the direction of motion from the reference image in inter-screen prediction.
In the intra prediction, since the prediction is performed from the pixels in the upper and left directions of the encoding target block, the direction of motion is limited as compared with the inter prediction.
When the inclination of the motion vector is far from the 33-direction directional prediction variations excluding the DC prediction of the intra prediction, it is considered that the directional prediction with the direction may not be suitable.
On the other hand, when the inclination of the motion vector is close to any of the variations of the directionality prediction, it is highly likely that the directionality prediction is suitable for the in-screen prediction.

Based on the description so far, the method for determining the intra prediction mode in the intra prediction mode determination unit 101 in the first embodiment will be described.
The procedure for determining the intra prediction mode is shown in FIG.

First, the intra-screen prediction mode determination unit 101 obtains the vector length of the input motion vector (S301).
The motion vector is input as coordinates to the reference image with the encoding target block as the origin, and the in-screen prediction mode determination unit 101 calculates the length from the coordinates.

  Next, the intra prediction mode determination unit 101 compares the vector length of the input motion vector with the magnitude of the vector length determination threshold value (S302).

When the vector length is larger than the vector length determination threshold by comparing the vector length with the vector length determination threshold (YES in S302), the intra prediction mode determination unit 101 has a certain amount of movement in a certain direction. It is determined that the block is a certain block, and the process proceeds to evaluation of the inclination determination threshold.
On the other hand, if it is less than the vector length determination threshold (NO in S302), as a block unsuitable directional intra prediction, intra prediction mode determination unit 101 selects the Plan a r prediction or DC prediction Then, the mode determination is finished (S303).

  Next, when the motion vector is longer than the threshold for vector length determination, the intra prediction mode determination unit 101 calculates the inclination of the motion vector and has a direction other than the DC prediction of the intra prediction. The difference absolute value between the inclination of the variation and the inclination of the motion vector is calculated, and the minimum value among the calculated difference absolute values is selected (S304).

Next, the in-screen prediction mode determination unit 101 compares the calculated minimum value of the absolute difference of inclination with the input inclination determination threshold (S305).
As a result of the comparison, if the minimum value of the absolute difference of the inclination is larger than the inclination determination threshold value (YES in S305), it is determined that it is not suitable for the prediction in the directionality prediction having the direction, and the in-screen prediction mode determination unit 101 ends the determination of the prediction screen by selecting Plan a r prediction or DC prediction.
When the minimum value of the absolute difference of the inclination is equal to or less than the inclination determination threshold (NO in S305), it is determined that the directionality prediction in the prediction direction is suitable for the intra prediction of the encoding target block, The intra prediction mode determination unit 101 selects the prediction mode in that direction as the intra prediction mode, and ends the determination of the intra prediction mode.

  As described above, in this embodiment, by determining the intra prediction mode without calculating the reference image of the intra prediction, and using the motion vector of the inter prediction while reducing the calculation related to the intra prediction. By capturing the characteristics of the block to be encoded, a moving image encoding apparatus that suppresses an increase in circuit scale and an increase in computation time while suppressing deterioration in image quality is realized.

As described above, in the present embodiment,
With a moving image as an input, an intra-screen prediction mode determination unit, an intra-screen prediction image generation unit, a local decoded image memory, a motion compensation prediction unit, a frame memory, a subtractor, a frequency conversion quantization unit, an inverse frequency conversion inverse quantization unit, A moving image encoding apparatus including an adder, an entropy encoding unit, an encoding control unit, and a deblocking filter has been described.

In the present embodiment,
The in-screen prediction mode determination unit inputs the motion vector of the encoding target block from the motion compensation prediction unit and the setting parameter from the outside, evaluates the motion vector, and predicts the in-screen based on the evaluated information and the setting parameter It has been described that the direction is determined and the prediction direction is output to the intra-screen prediction image generation unit.

In the present embodiment,
The intra-screen prediction image generation unit inputs the intra-screen prediction mode determined by the intra-screen prediction mode determination unit and the local decoded pixels around the encoding target block necessary for predictive image generation from the local decoded image memory, and supports the prediction mode The generation of the predicted image within the screen and outputting it were explained.

In the present embodiment,
The local decoded image memory inputs the peripheral image of the encoding target block from the adder and the intra-screen prediction image generation unit, accumulates pixels necessary for generating the intra-screen prediction image, and outputs the pixels to the intra-screen prediction image generation unit. Was explained.

In the present embodiment,
The motion compensation prediction unit inputs an original image of the encoding target block and an image for reference from the frame memory, compares the encoding target block with the reference image, and determines the reference image having the highest similarity from the frame memory. It has been described that the extracted inter-frame prediction image is output to the selector, and the motion vector of the position information with the highest similarity of the reference image is output to the intra prediction mode determination unit and the encoding control unit.

In the present embodiment,
The frame memory inputs the locally decoded image subjected to the deblocking filter processing from the deblocking filter unit, accumulates the pixels necessary for generating the inter-frame prediction image, and based on the position information based on the motion vector input from the motion compensation prediction unit. In addition, it has been described that the pixels necessary for generating the inter-frame prediction are output to the motion compensation prediction unit.

In the present embodiment,
Explaining that an adder inputs a pixel difference value obtained by local decoding of an encoding target block from an inverse frequency transform inverse quantization unit and a prediction image from a selector, adds them, and outputs them to a deblocking filter unit and a local decoded image memory did.

In the present embodiment,
It has been described that the subtracter inputs the encoding target block and the prediction image of the encoding target block, subtracts them, and outputs them to the frequency transform quantization unit.

In the present embodiment,
The frequency conversion quantization unit inputs the pixel difference value from the subtractor, performs frequency conversion operations such as direct conversion and discrete cosine transform (DCT), and then quantizes the operation results to quantize the frequency. It has been described that the transform coefficient is output to the entropy coding unit and the inverse frequency transform inverse quantization unit.

In the present embodiment,
The inverse frequency transform inverse quantizer inputs the frequency transform coefficient quantized from the frequency transform quantizer, performs inverse frequency transform on the coefficient, performs inverse frequency transform operation, and decodes the pixel difference The output of the value to the adder has been described.

In the present embodiment,
It has been described that the entropy encoding unit inputs the frequency transform coefficient after quantization, performs entropy encoding, and outputs the encoded stream.

In the present embodiment,
It has been described that the encoding control unit controls the entire encoding.

In the present embodiment,
It has been described that the deblocking filter unit performs a deblocking filter operation on the locally decoded image input from the adder and outputs the result to the frame memory.

Embodiment 2. FIG.
FIG. 4 shows a configuration example of the moving picture coding apparatus 100 according to the second embodiment.
In FIG. 4, an image feature amount extraction unit 114 is added as compared to the configuration of FIG.
The image feature amount extraction unit 114 receives the image of the encoding target block, calculates the variance of the pixel values of the pixels included in the encoding target block, and outputs the obtained variance to the in-screen prediction mode determination unit 101.
Elements other than the image feature quantity extraction unit 114 are the same as those shown in FIG.
In the present embodiment, points different from the first embodiment will be described, and description of the same points as in the first embodiment will be omitted.

  The operation will be described below.

The intra-screen prediction mode determination unit 101 inputs parameters for setting from the outside, and receives a motion vector derived from the motion compensation prediction unit 104 for inter-frame prediction of the encoding target block.
Specifically, the setting parameters are a threshold for vector length evaluation, a threshold for vector inclination evaluation, and a threshold for block distribution evaluation encoding, and does not use a reference image between frames. Then, a motion vector corresponding to inter-frame prediction is also derived for a picture to be predicted in the screen, and the motion vector is input to the intra-screen prediction mode determination unit 101 for determining the intra-screen prediction mode.

Since the evaluation of the motion vector with respect to the threshold for vector length evaluation and the evaluation of the motion vector with respect to the threshold for vector inclination evaluation are the same as those in the first embodiment, a description thereof will be omitted.
The evaluation of the encoding target block with respect to the encoding target block variance evaluation threshold will be described.
If the variance value of the encoding target block is smaller is the variation is small block is flat area of the block in the pixel value, it is likely that suitable is DC prediction when comparing DC prediction and Plan a r prediction.
Conversely, when the variation is large block is not a flat region, it is likely that suitable is Plan a r prediction when comparing DC prediction and Plan a r prediction.

Based on the description so far, the method for determining the intra prediction mode in the intra prediction mode determination unit 101 in the second embodiment will be described.
The procedure for determining the intra prediction mode is shown in FIG.

In FIG. 5, S501, S502, S507, S508, and S509 are the same as S301, S302, S304, S305, and S306 in FIG.
Below, operation | movement of S503-S506 is demonstrated.

When the vector length is equal to or less than the vector length determination threshold value (No in S502) and when the minimum absolute value of the gradient difference is larger than the gradient determination threshold value (Yes in S508), the intra prediction mode The determination unit 101 causes the image feature amount extraction unit 114 to calculate the variance value of the encoding target block (S503).
Then, the intra prediction mode determination unit 101 compares the variance determination threshold value with the calculated variance value of the encoding target block (S504).
As for the subsequent intra prediction mode selection, since it has already been determined to be unsuitable directional prediction with direction and selecting from the DC prediction and Plan a r prediction.
As a result of the comparison, if the variance value of the encoding target block is smaller than the variance determination threshold value (YES in S504), it is determined that the encoding target block is a flat region with small pixel variation, and intra prediction The mode determination unit 101 selects DC prediction (S505).
On the other hand, when the variance value of the encoding target block is larger than the variance determination threshold value (NO in S504), it is determined that the encoding target block is not a flat region with relatively large pixel variation, and the intra prediction mode The determination unit 101 selects the Plan a prediction (S506).

  As described above, in this embodiment, by determining the intra prediction mode without calculating the reference image of the intra prediction, and using the motion vector of the inter prediction while reducing the calculation related to the intra prediction. By capturing the characteristics of the block to be encoded, a moving image encoding apparatus that suppresses an increase in circuit scale and an increase in computation time while suppressing deterioration in image quality is realized.

Embodiment 3 FIG.
FIG. 6 shows a configuration example of the moving picture encoding apparatus according to the third embodiment.
In FIG. 6, a peripheral motion vector holding unit 115 is added as compared to the configuration of FIG.
The peripheral motion vector holding unit 115 receives the motion vector from the motion compensated prediction unit 104, holds the motion vectors of the blocks around the encoding target block, and stores the motion vector of the peripheral block of the encoding target block in the intra prediction mode determination unit 101. Output motion vector.
Elements other than the peripheral motion vector holding unit 115 are the same as those shown in FIG.
In the present embodiment, points different from the first embodiment will be described, and description of the same points as in the first embodiment will be omitted.

  The operation will be described below.

The intra-screen prediction mode determination unit 101 receives a setting parameter and a motion vector derived from the motion compensation prediction unit 104 for inter-frame prediction of the encoding target block from the outside, and from the peripheral motion vector holding unit 115 The motion vector of the block around the encoding target block is input.
Specifically, the setting parameters are a threshold for vector length evaluation, a threshold for vector inclination evaluation, and a threshold for peripheral motion vector evaluation, and does not use a reference image between frames. Then, a motion vector corresponding to inter-frame prediction is also derived for a picture to be predicted in the screen, and the motion vector is input to the intra-screen prediction mode determination unit 101 for determining the intra-screen prediction mode.

FIG. 7 shows the procedure for determining the intra prediction mode in the third embodiment.
S704 to S708 are the same as S301 to S306 in FIG.

  In the procedure of the third embodiment, first, in order to evaluate the variation between the motion vector of the peripheral block to be encoded and the motion vector of the block to be encoded input from the peripheral motion vector holding unit 115, the intra prediction mode determination is performed. The unit 101 calculates a variance for the difference between the motion vector of the encoding target block and the motion vectors of the neighboring blocks (S701).

If the calculated variance value is larger than the threshold for evaluating the surrounding vector (YES in S702), the intra prediction mode determination unit 101 determines that it is difficult to estimate the encoding target prediction direction. and, to end the DC prediction or Plan a r decision to select the prediction intra prediction mode.
In other cases (NO in S702), since there is a possibility of being suitable for directionality prediction having a direction, the intra-screen prediction mode determination unit 101 proceeds to the procedure after calculating the vector length of the motion vector.

  As described above, in this embodiment, by determining the intra prediction mode without calculating the reference image of the intra prediction, and using the motion vector of the inter prediction while reducing the calculation related to the intra prediction. By capturing the characteristics of the block to be encoded, a moving image encoding apparatus that suppresses an increase in circuit scale and an increase in computation time while suppressing deterioration in image quality is realized.

As mentioned above, although embodiment of this invention was described, you may implement in combination of 2 or more among these embodiment.
Alternatively, one of these embodiments may be partially implemented.
Alternatively, two or more of these embodiments may be partially combined.
In addition, this invention is not limited to these embodiment, A various change is possible as needed.

  DESCRIPTION OF SYMBOLS 100 moving image encoder, 101 intra prediction mode determination part, 102 intra prediction image generation part, 103 local decoding image memory, 104 motion compensation prediction part, 105 frame memory, 106 adder, 107 subtractor, 108 frequency conversion Quantization unit, 109 inverse frequency transform inverse quantization unit, 110 entropy encoding unit, 111 encoding control unit, 112 deblocking filter unit, 113 selector, 114 image feature quantity extraction unit, 115 peripheral motion vector holding unit.

Claims (7)

A video encoding device that performs inter-frame prediction and intra-frame prediction,
For each image block divided from the image frames constituting a moving image, enter the motion vector calculated for the inter-frame prediction, and calculates the vector length of an input motion vector for each image block, the vector calculated When the calculated vector length is equal to or less than the threshold for vector length determination , the Planar prediction mode and the DC prediction are used as the in-screen prediction modes applied to the in-screen prediction. A moving picture coding apparatus comprising: an in-screen prediction mode determination unit that selects one of modes . The in-screen prediction mode determination unit
When the calculated vector length is larger than the vector length determination threshold,
Calculate the vector slope of the input motion vector,
And vector tilt issued calculated, the difference absolute value of the gradient associated with the each of the plurality of directional prediction modes except for DC prediction mode is calculated,
Select the minimum difference absolute value from the calculated multiple difference absolute values,
Compare the selected absolute difference with the threshold value for slope judgment,
If the minimum difference absolute value chosen is greater than the determining threshold value gradient, select one of the Plan a r prediction mode and the DC prediction mode,
If the minimum difference absolute value selected is equal to or less than the determining threshold value slope, minimum video according to claim 1, the absolute value difference and selects a directional prediction mode as an object selected Image encoding device. The moving image encoding device further includes:
The image prediction mode determination unit includes an image feature amount extraction unit that calculates a variance value of pixel values of pixels included in an image block that is a target of selection of the image prediction mode;
The in-screen prediction mode determination unit
When the calculated vector length is less than or equal to the vector length determination threshold,
The image feature amount extraction unit calculates a variance value of pixel values of pixels included in an image block that is a target of selection in the intra prediction mode, and inputs a variance value from the image feature amount extraction unit,
Compare the entered variance value with the threshold value for variance judgment,
When the input variance value is smaller than the threshold value for variance determination, the DC prediction mode is selected,
When entered variance value is greater than or equal variance determining threshold value, Plan a r a moving image encoding apparatus according to claim 1 or 2, characterized in that selects a prediction mode. The moving image encoding device further includes:
The image prediction mode determination unit includes an image feature amount extraction unit that calculates a variance value of pixel values of pixels included in an image block that is a target of selection of the image prediction mode;
The in-screen prediction mode determination unit
When the selected minimum difference absolute value is larger than the threshold value for slope judgment,
The image feature amount extraction unit calculates a variance value of pixel values of pixels included in an image block that is a target of selection in the intra prediction mode, and inputs a variance value from the image feature amount extraction unit,
Compare the entered variance value with the threshold value for variance judgment,
When the input variance value is smaller than the threshold value for variance determination, the DC prediction mode is selected,
The moving image encoding apparatus according to claim 2 , wherein the Plan a r prediction mode is selected when the input variance value is equal to or greater than a variance determination threshold value. The moving image encoding device further includes:
The intra prediction mode determination unit includes a peripheral motion vector holding unit that holds a motion vector calculated for an image block located in the periphery of the image block to be selected in the intra prediction mode;
The in-screen prediction mode determination unit
From the peripheral motion vector holding unit, input the motion vector of the image block located in the periphery of the image block to be selected in the intra prediction mode,
Calculating the difference between the image block to be selected in the in-screen prediction mode and the motion vector input from the peripheral motion vector holding unit, calculating a variance value in the calculated difference;
Compare the calculated variance with the threshold for evaluating the surrounding vector,
If the calculated variance value is larger than the threshold for evaluating the surrounding vector, select one of the Plan a prediction mode and the DC prediction mode,
When the calculated variance is less than or equal to the threshold for evaluating the surrounding vector,
Calculating a vector length of the motion vector of the image block of interest in the intra prediction mode selection, one of the claims 1 to 4, characterized in that comparing the calculated and the vector length and vector length determining threshold value A moving image encoding apparatus according to claim 1. An intra-screen prediction mode determination method by a computer in which inter-frame prediction and intra-screen prediction for a moving image are performed,
The computer inputs the motion vector calculated for inter-frame prediction for each image block divided from the image frames constituting the moving image, calculates the vector length of the input motion vector , and for each image block When the calculated vector length is compared with the vector length determination threshold and the calculated vector length is equal to or less than the vector length determination threshold , Planar prediction is used as the intra prediction mode applied to the intra prediction. A method for determining an in-screen prediction mode, wherein one of a mode and a DC prediction mode is selected.   A program for causing a computer to function as the moving image encoding apparatus according to claim 1.

Images